Characterization of Irradiated Doping Profiles. Wolfgang Treberspurg, Thomas Bergauer, Marko Dragicevic, Manfred Krammer, Manfred Valentan

Transcription

1 Characterization of Irradiated Doping Profiles, Thomas Bergauer, Marko Dragicevic, Manfred Krammer, Manfred Valentan Vienna Conference on Instrumentation (VCI)

2 Content: Experimental Procedure Sample Irradiation Spreading Resistance Profiling (SRP) Theoretical Background The Electrical Neutral Bulk (ENB) The Space Charge Region (SCR) Results Irradiated Profiles of Shallow Impurities Effective Doping Concentration References

3 3 Experimental Procedure

4 4 Experimental Procedure: Irradiation Samples: Floating-zone, p-on-n diodes 1 kωcm bulk resistivity Produced by HPK Bulk and backside implant doped with phosphor Overview: Sample irradiation Annealing of 8 minutes at 60 C Mechanically cut Preparation for SRP measurements Common capacitance voltage measurements Sample Selection Irradiation and Annealing SRP Measurements CV Measurements

5 5 Experimental Procedure: Irradiation Irradiation facility: TRIGA Mark-II nuclear reactor At the Institute of Atomic and Subatomic Physics (ATI) Vienna Previous fluence calibration with similar diodes 27 C during irradiation place Spectrum approximately known

6 6 Experimental Procedure: Irradiation Calculation: Hardness factor κ has been measured during calibration Calculation of κ confirmed calibration results Damage mainly caused by thermal and fast neutrons D( E) ( E) de D(1MeV ) ( E) de V I Sample Φ eq [n eq /cm 2 ]

7 Experimental Procedure: SRP Preparation: The samples are mechanically grinded and polished with a small bevel angle During preparation no temperatures above 30 C are expected Measurement: Two closely aligned probes are stepped along the beveled surface The probes are contacted with reproducible weight The probes are made of hard material (tungsten carbide) The resistance is measured at each position

8 8 Experimental Procedure: SRP Data Extraction: The measured resistance is dominated by current spreading effects (R sp ) Very small radii (a) of the probe tips result in a high sensitivity on the material resistivity Depth resolution of about 100 nm can be easily achieved (10 µm planar) SRP measurements deliver no absolute data, the accuracy depends on the calibration effort Only electrically active dopants contribute to the measurement Performance: The measurements have been done in a custom designed setup [1] R sp 2a ( x) ( qn( x) ) 1

9 9 Theoretical Background

10 10 Theoretical Background: ENB The electrical neutral bulk Non depleted silicon volume At sensor operated with voltages smaller than the full depletion voltage The ionization of traps is determined by the Fermi level occupancy function at equilibrium Suited for direct resistivity measurements Modelling of resistivity The shift of the fermi level (E F ) after irradiation is determined by the charge neutrality equation Ionization of acceptor N a - and donor traps N d + has to be taken into account The resistivity is determined by concentration of free charge carrier inside the ENB (n,p) Models have to assumed the number and energy level of traps [2,3] SCR ENB n N a p N d

11 11 Theoretical Background: SCR The space charge region Inside the SCR the ionization of deep traps is regulated by band bending in respect to the potential distribution Almost all deep traps contribute to the effective doping concentration N eff Donor removal processes Vacancies or self interstitials remove dopants from regular sides Donor removal takes place until a non removable amount of donors is reached, which depends on the presence of other elements inside the bulk [4] In turn the donor removal constant c also depends on the initial concentration N eff,0

12 12 Theoretical Background: SCR Capacitance measurements at the presents of deep traps Deep trap usually only contribute at low frequencies, in respect to their time constant As the dc bias voltage correlates to very small frequencies most traps inside the SCR (X Trap ) are ionized Only traps inside the volume dx Ampl which is depleted by the ac test signal are sensitive on the measurement frequency [5] Example The contribution to capacitance measurements of one deep donor like trap is shown Strong influence of measurement parameters: Frequency (responding traps) Amplitude (sensitive volume) Temperature (time constant changes with T) Measurement mode (serial or parallel)

13 13 Measurement Results

14 14 Results: Shallow Impurities Profiles after irradiation Due to donor removal the expected resistivity decrease is observed The donor removal process results in an decreasing implantation depth for increasing fluences Explanation Uniformly distributed defects are expected to shift the profile only Shifting a profile results in a linear correlation of the initial concentration to the resulting one Effect can be understood by assuming that the resulting bulk concentrations converges to the non removable donor concentration Another approach is to assume spatial effects due to high resistivity gradients

15 15 Results: Shallow Impurities Annealing study To investigate a possible influence of annealing on this effect several subsequent annealing steps have been performed (0, 8, 28, 120 minutes at 60 C) No further changes of profiles could be observed, the effect has to take place during irradiation Only shallow impurities are measured, especially after high fluences the electrical behaviour is dominated by deep traps

16 16 Results: Effective Doping Concentration Electrical behaviour The influence of the observed effect on the electrical performance is studied by using capacitance voltage measurements Samples irradiated with more than last ones type inverted n eq /cm 2 have been observed to be type inverted Capacitance measurements become strongly frequency dependent, which make it advisable to use capacitance voltage frequency surfaces C-V Measurement: Parallel mode, -15 C, Amplitude: 200mV, Frequency: 1k

17 Results: Effective Doping Concentration 1E11 5E11 2E12 3E12 1E13 5E13 1E14 5E14 C-V Measurements on diodes irradiated with increasing fluencies: Parallel mode, -15 C Amplitude: 200mV

18 18 Results: Effective Doping Concentration Electrical behaviour The influence of the observed effect on the electrical performance is studied with capacitance voltage measurements Type inverted samples feature a peak at low frequencies due to dominant deep traps The capacitance stabilize at high frequencies for all samples at a value associated with the geometrical capacitance [6] High fluences result in an decreasing dependence on the bias voltage C-V Measurement: Parallel mode, -15 C, Amplitude: 200mV, Frequency: 1k, Bias Voltage: 50V

19 19 Results: Effective Doping Concentration Doping profiles of N eff By differentiating C -2 a profile of the effective doping concentration inside the SCR is calculated [7] At the presented low fluence range donor removal effects are dominant and the concentration decreases With increasing fluences the profiles become inhomogeneous, especially close to the backside, where donor removal mainly takes place C-V Measurement: Parallel mode, - 15 C, Amplitude: 200mV, Frequency: 1k

20 20 Conclusion and Outlook Conclusions It could be shown, that SRP measurements are suited to characterize profiles of shallow donors after irradiation An effect of decreasing implantation depth of shallow impurities with increasing fluences could be observed To evaluate the influence of this effect on the sensor performance capacitance voltage measurements are suited Outlook Capacitance voltage measurements have to be investigated in future and may serve for more detailed impurity characterization The measurement of shallow impurities only with high spatial resolution may be used for investigating their correlation to other impurities

21 21 References [1] W. Treberspurg et al., Measuring Doping Profiles of Silicon Detectors with a Custom-designed Probe Station, Journal of Instrumentation 7 (2012). /doi: / /7/11/p11009 [2] E. Borchi et al., Radiation Damage in Silicon Detectors, La Rivista del Nuovo Cimento, vol. 17, (1994) p [3] M. Bruzzi, Radiation Damage in Silicon Detectors for High-Energy Physics Experiments, IEEE Transactions on nuclear science 48/4 (2001) 960. /doi: / [4] M. Moll et al., Investigation on the improved radiation hardness of silicon detectors with high oxygen concentration, Nucl. Instr. and Meth. A 439 (2000). /doi: /s (99) [5] W. Dabrowski et al., Effects of deep imperfection levels on the capacitance of semiconductor Detectors, Nucl. Instr. and Meth. A 276 (1989). [6] Z. Li et al., Studies of Frequency Dependent C-V Characteristics of Neutron Irradiated p-n Silicon Detectors, IEEE Transactions on nuclear science, Vol. 38, No. 2, (1991) [8] j. Shiau et al., Interpretation of capacitance versus voltage measurements in the presence of a high density of deep levels, J. Appl. Phys. 59, 2879 (1986); doi: /